Sanofi-Aventis Seeks the Holy Grail of Insulin

by: Joseph Krueger
Last week, Sanofi-Aventis (NYSE:SNY) confirmed that it is on the prowl for diabetes partnerships. Bloomberg reported statements made by Riccardo Perfetti, the Senior Medical Director for Sanofi-Aventis’ diabetes unit, that Sanofi-Aventis may unveil “at least another couple critical partnerships” in diabetes in the months to come. Most interesting to me was a quote stating that they: “…may consider candidates for a successor to its best-selling product, the long-acting insulin Lantus, and for a new, fast-acting insulin."
Given my recent interest in fast acting insulins up for approval soon, I spent some time considering this statement: Who would Sanofi-Aventis most likely be after?
I came up with three novel fast-acting insulin delivery systems which could be up for potential partnership: Mannkind’s (NASDAQ:MNKD) Afrezza, Biodel’s (BIOD) Linjeta, and Halozyme’s (NASDAQ:HALO) rHuPH20 insulin.
These three candidates represent unique approaches to insulin delivery, and are all strong candidates to fill the needs which remain in the fast-acting insulin market. As an investor in this arena, I clearly would like to know which one is the most likely one up for partnership with SNY. In order to make a prediction, investors need to consider what needs these candidates could fill for SNY. To make this assessment, investors need to have some background about the existing fast acting insulin market, and the “next generation” players who are also seeking partnerships. So first, some background:
Types of existing fast acting insulins:
The existing fast acting insulins all work on the same principle: The insulin is changed by altering its structure (substituting amino acids) so that it does not form hexamers. Native insulin is found in the hexameric state, which decreases its absorption rate when injected. Herein lies the problem: diabetics trying to time their insulin dose to control blood sugar find that it takes too long for regular insulin to enter the blood stream, causing blood sugar spikes (hyperglycemia); and due to its slow absorption remains in the blood stream too long, causing dangerous blood sugar dips (hypoglycemia). Thus, the most critical factor for insulins are their pharmacokinetic profile; how quickly they enter and remain in the bloodstream.

The fast-acting analogues of insulin are altered to have charge repulsion at key locations on the molecule, which prevents the formation of insulin hexamers, and decreases the time it takes for them to enter the blood stream.
For the scientists, Insulin Aspart (Novolog) is regular insulin which has been mutated so that the amino acid proline on the beta-chain of insulin (B28), is substituted with an aspartic acid residue. Insulin lispro (Humalog) is made by switching positions B28 (Lys) and B29 (Pro) on the insulin beta -chain. Insulin glulisine (Apidra) is made by replacing the amino acid asparagine in the beta chain at position B3 with lysine and the lysine in position B29 with glutamic acid.
For the layman, all of these three modified insulins work in very similar ways: 1-2 mutations at critical sites in the insulin molecule prevent hexamers from forming, and the insulin is absorbed quicker.
Due to the similar nature of these insulins, all three of these fast acting insulins have nearly identical pharmacokinetic profiles (insulin lispro vs insulin glulisine shown below). These fact acting insulins distinguish themselves from standard recombinant human insulin (RHI), but not significantly apart from each other.

The major fast acting insulin players:
Of the three main insulin producers, Eli Lilly (NYSE:LLY), Sanof- Aventis and Novo Norodisk (NYSE:NVO): Novo Nordisk has been involved in the diabetes market since insulin was first developed in the 1920s. Novo Nordisk sales of fast-acting insulin Novolog (insulin aspart) is the longest-established and highest-selling product in the fast acting insulin market. Eli Lilly is the second largest competitor with its fast-acting insulin product Humulog (insulin lispro). Both of these companies are stiff competition for Sanofi-Aventis, of the three the newest comer to the insulin market with Apidra (insulin glulisine). Sanofi has taken the lead of the long-acting insulin analog market with its blockbuster product, once-daily Lantus (insulin glargine), but Apidra sales pale in comparison to Novalog and Humalog.

Sanofi-Aventis’s position in the insulin market:
SNY’s global Lantus sales are projected to grow to an astounding $6 billion by 2014 when the patent expires, but SNY is struggling to compete in the fast-acting insulin market with Apidra due to its competition. Apidra wasn’t launched in the US until 2009, so sales have been minor compared to the competitors. However, Apidra was launched in the EU in 2005 and sales in 2009 were about $300 million. Compared to Novolog sales of $900 million and Humalog sales of $750 million in 2009, Apidra is likely to only have a minor share of the US fast acting insulin market. However, the US patent for Novalog expires next year, and the US patent for Humalog expires in 2014. Given that there is no fundamental or clinical basis for differentiating these three fast acting insulins (as described above), continued market pressure from the brand name and future generic forms of these drugs are likely to put sustained pressure on Apidra sales, even though Apidra has US patent protection until 2018. In consideration of these factors, and a loss of Lantus patent protection in 2014, Sanofi-Aventis is logically looking for a way to differentiate itself from the other fast acting insulins; as well as find a product that can replace Lantus sales by 2014.

This is the clear basis for the statement that they “may consider candidates for a successor to its best-selling product, the long-acting insulin Lantus, and for a new, fast-acting insulin. Given the competitive state of market, the equivalent nature of the three fast acting insulins available, the time frame in which these players lose market exclusivity, and Sanofi’s position in this market, it makes perfect sense that Sanofi is looking for a new fast-acting insulin partnership and a solution to protect Lantus.
The “next generation” insulins
Novel insulins, like Mannkind’s Afrezza, Halozyme’s rHuPH20 insulin, and Biodel’s Linjeta have the same goal of rapid control of glucose, but take a vastly different approach than Apidra, Humalog, or Novalog. As well, these three companies all take vastly different approaches from each other. Afrezza’s approach is to utilize chemically modified insulin which remains as monomers (Technosphere) and the dense lung vasculature to absorb inhaled insulin quickly. Linjeta’s approach is also to use chemically modified insulin to prevent insulin hexamers and thus fast absorption by injection. Halozyme’s approach is to change the tissue at the site of injection to allow faster absorption. All these approaches work better than the existing insulins Apidra, Novolog, and Humalog as demonstrated below:
When compared directly to insulin lispro, Afrezza is absorbed very quickly into the blood, reaching peak plasma concentrations in 15-20 minutes; compared to 50-60 minutes for insulin lispro.

When compared directly to insulin lispro, Linjeta is also absorbed very quickly into the blood, reaching peak plasma concentrations by 25-30 minutes; compared to 50-60 minutes for insulin lispro.

The concept behind Halozyme’s rHuPH20 insulin is different than both Afrezza and Linjeta. The concept behind Halozyme’s “Enhanze” technology is the use of an enzyme called hyalronidase that locally degrades body components under the skin to enable quicker absorption of a drug (like insulin) into the blood. The principle behind this is that the layer of tissue under skin is composed partly of hyalronic acid, a water-absorbing protein that gives tissues under the skin their “bounce”. Halozyme’s Hylenex hyalronidase degrades the hyalronic acid locally at the site of injection, disrupting tissue structure, and allowing absorption into neighboring capillaries quicker. Halozyme has gained approval for their Hylenex (rHuPH20) drug with co-administration with other drugs for certain pediatric uses.
In the case of this drug, there is still the problem of insulin forming hexamers, so a fast acting insulin such as Novolog, Humalog, or Apidra must still be used. However, when combined, there is quicker absorption of the insulin into the blood stream as shown below. This does not appear to be faster than Afrezza or Linjeta, but is certainly an improvement over insulin lispro.

Humalog and Novolog have improved on their stand-alone fast acting insulin products by marketing mixes of fast-acting and regular insulin (Novolog 70/30, and Humalog 70/30). However, these mixes are only for diabetics who need insulin in between mealtime as well as at mealtime. The basal insulin in these products does not last as long as the long-lasting Lantus, and interim injections of basal insulin are often needed despite this mixture. Thus, there are two gaps to fill: First, better mealtime control of insulin that Afrezza, Linjeta offers, and potentially Halozyme’s rHuPH20 insulin can offer; and second an all-inclusive insulin that only requires use at mealtime and provides coverage between meals. Solving these problems are the “holy grail” of insulin, and any product brought to market which offers this clearly has a competitive advantage over all existing and future products.

Who will SNY choose to partner with?
None of these products is mutually exclusive; SNY inferred multiple partnerships and may choose one or all (or none) of these as partners, as each one has unique characteristics and thus SNY would have unique rationale for choosing each one. However, because of these unique characteristics, we can weigh what may be the most likely partner for SNY given the current fast acting insulin environment.
Formulating the fast acting insulin with Halozyme’s rHuPH20 may in fact give added benefit to diabetics (improved glucose control, less hypoglycemia, weight loss) due to improve pharmacokinetics, but this remains to be proven. It is clear that Linjeta and Afrezza have distinct advantages over the current fast acting insulins; if Novolog, Humalog, and Apidra could improve their pharmacokinetic profile by using rHuPH20, they would be able to compete with them. Also, Halozyme’s formulation offers means for additional patents and market exclusivity extension for the fast acting insulins facing patent expiration relatively soon (Novolog, Humalog).
However, Halozyme has only completed studies with insulin lispro. They have just begun their phase II studies with Humalog and Novolog, which could take a couple years, and if successful would put them up for FDA approval sometime in 2014. This would be too late for Novolog (2011); and right on time for Humalog (2014). The fact that they are using lispro insulin, and the timing needed to complete trials, positions them with a partnership with Eli Lilly to use with Humalog. The intended use and timing seems wrong for a HALO-SNY partnership and best for a HALO-LLY partnership.
In contrast, the timing for Mannkind’s Afrezza is excellent. Certainly Afrezza meets SNY’s desire for an improved fast acting insulin product. After several delays, Afrezza is up for approval again on December 29th of this year. This seems to fit the time frame for partnership SNY is seeking. However, Afrezza is a novel drug/device combination, which requires the 505(b)1 approval route and the combination drug/device is making gaining FDA approval more difficult ( this is the cause for repeated delays in its approval). If delayed again this December, it would still be up for approval again in early 2011, which seems to still fall inside of the time frame SNY may announce a partnership.
However, as I discussed in a previous article, inhaled insulin carries significant marketing acceptance. Inhaled insulin also may require risky post-labeling safety risks due to real or perceived potential pulmonary risks associated with inhaled insulin (black box warning) and post-marketing follow-up studies. In addition to this, a partnership with MNKD would have to be quite expensive to start to recoup the well over $1 billion development costs of Afrezza. Mannkind has maintained that Afrezza would be competitive in pricing to existing fast acting insulins, which makes profit margins lower and harder to justify and an expensive partnership with MNKD for a good return on investment. These are often stated reasons as to why Mannkind has not come forward with a partnership yet, and may also be reasons why SNY might be hesitant to partner Afrezza.
Yet, this is not reason alone to exclude a SNY-MNKD partnership. As well, a MNKD-NVO partnership seems just as likely given the urgency required for NVO to replace lost Novolog sales due to loss of marketing exclusivity in 2011; or similarly a LLY-MNKD partnership to replace Humalog (although neither company has stated they are seeking partners). But one major problem with Afrezza in relation to the strategy of such partnerships is that it does not fulfill the need to replace the Novalog 70/30 and Humalog 70/30 fast-acting/basal insulin combinations, as Afrezza is inhaled and basal insulin is injected. Thus, although Afrezza may improve fact-acting insulin, it cannot be used immediately to replace these products, nor to develop the “holy grail” combination of a singular fast-acting/long lasting insulin product.
However, Linjeta is perfectly positions to accomplish this. As I wrote in a previous article, the injectable insulin platform is one that is comfortable to drug companies (and insulin users), and seemingly provides the least amount of marketing risk. In addition to this, injectable insulins pose little pre-and-post marketing regulatory risk, as they have been used for decades. Biodel’s Linjeta stands out in this arena not only as a superior fast acting insulin over the existing marketed products, but is seeking approval under the 505(b)2 mechanism as a new formulation of an existing drug. Apart from the troublesome Indian data (discussed here), there are no clear risks for approval, and no perceived risked for marketing. Linjeta would easily fill the need for an improved fast-acting insulin to gain a decent market share in the face of Apidra’s existing competitors Novolog and Humalog; as well as the generic versions that will be on the market before Apidra loses market exclusivity.
It is easy to rationalize why SNY may favor an injectable insulin over Afrezza, as the marketing would be the same as Apidra but SNY could claim superiority over Novolog and Humalog. BIOD plans to market Linjeta in the U-100 (100 units insulin/ml) formulation in 10ml vials or 3ml injector pens- identical to the current Apidra products sold as U-100 formulation in 10ml vials or 3ml cartridges for use with an insulin pen marketed as Apidra SoloSTAR. This format is also available to Novolog and Humalog users, so it is easy to see how integration of an improved insulin into an existing use marketing space would be easy and successful. Thus, it seems more likely that SNY would choose this format (and thus Linjeta) over the inhaled insulin (Afrezza) format for partnership.
But there are more pieces to the puzzle that heavily sways bias towards SNY choosing BIOD as a partner over all others: Lantus. The long acting insulin Lantus fills a need not covered by fast acting insulins for diabetics who need insulin between meals as well as at mealtime. A considerable portion of Lantus sales also comes from its use instead of insulin pumps. However, with Lantus losing marketing exclusivity in 2014, SNY needs to keep its share of this market by introducing a new, superior product to the market. BIOD is currently testing the use of Linjeta in insulin pumps (trial NCT01067118), and would fill this need (if successful) by 2014 when Lantus comes off patent.

However, one of the key pieces of evidence, which makes me believe that BIOD would be the partner of choice for SNY, is the need for SNY to bring to market an improved long acting/ fast acting injectable combination mixture to outcompete all the others (the “holy grail”). As previously mentioned, NVO sells Novolog 70/30, and LLY sells Humalog 70/30; fast acting insulin/basal insulin mixtures for the purpose of providing glucose control at meals and between meals. However, SNY does not sell an Apidra/Lantus mixture because Lantus cannot be mixed with mixed with Apidra or any other fast acting insulin as the mixture causes Lantus to precipitate prior to injection and administration, and this makes it virtually impossible to administer a known and reliable dose.
Fortunately for SNY, someone has designed a unique product specifically to fill in this gap: Investors should note that Biodel has quietly gained US patent 11695562 pertaining the use of fast and long acting insulin mixtures for injection. The patent pertains specifically to the use of injectable mixture of glargine insulin (Lantus), and VIAject/Linjeta.
Linjeta (VIAject) is a unique fast-acting insulin compatible with Lantus due to its low pH formulation. Although stand-alone Linjeta has been reformulated to a neutral pH for comfort, the phase III Linjeta trials were performed using the more acidic formulation. This appeared to be a boondoggle for BIOD, as the pH caused some discomfort at the injection site; but really appears to be a boon for BIOD as this formulation allows it to be combined directly with Lantus, as the patent demonstrates. Given that the pH of the Linjeta formulation can be changed for different uses, but Lantus cannot, this positions Linjeta as the perfect drug to be combined with Lantus for diabetics who need both short term blood sugar control at mealtime and long term insulin between meals (the “holy grail”). As well, Linjeta can be used separately as the to-be-marketed neutral pH formulation in injection pens at mealtime by diabetics who do not need long acting insulin between meals. However, Linjeta is the only product with the right chemistry that fulfills both of the needs SNY is looking to fill, and has already obtained the intellectual property specifically surrounding this.

BIOD has quietly developed an improved mixed insulin product compatible with Lantus, which would allow the “holy grail” combination fast-acting/long acting insulin to be created. This product, if created, fills the current unmet need of fast/long acting insulin mixtures, and also fills the exact niche that SNY needs to corner the market. This partnership alone would allow patent extension for Lantus past 2014 as the Lantus/Linjeta mixture (say it out loud - is this why they changed the name from VIAject to Linjeta?). This partnership would also provide an improved fast-acting insulin that can compete with Novalog and Humalog sales, and their eventual generic counterparts as it provides superior benefits.

The quote made by the SNY spokesperson that they “may consider candidates for a successor to its best-selling product, the long-acting insulin Lantus, and for a new, fast-acting insulin” basically spells this out in plain words. In my opinion, this partnership makes too much sense to be discounted. The idea of this partnership is not just conjecture; the patent for the Lantus/Linjeta mixture is an obvious business strategy that clearly BIOD and presumably SNY have been considering for quite some time now.

Disclosure: Long BIOD